SCIENTIFIC OPINION. EFSA Panel on Genetically Modified Organisms (GMO) 2,3. European Food Safety Authority (EFSA), Parma, Italy

Transcription

1 EFSA Journal 2014;12(5):3644 SCIENTIFIC OPINION Scientific Opinion on application (EFSA-GMO-UK ) for the placing on the market of soybean MON genetically modified to contain stearidonic acid, for food and feed uses, import and processing under Regulation (EC) No 1829/2003 from Monsanto 1 ABSTRACT EFSA Panel on Genetically Modified Organisms (GMO) 2,3 European Food Safety Authority (EFSA), Parma, Italy Soybean MON was developed using Agrobacterium tumefaciens transformation and was intended to modify the lipid profile of the extracted oil. Soybean MON contains a single insert consisting of the Pj.D6D gene encoding the Δ6 desaturase protein from Primula juliae and the Nc.Fad3 gene encoding the Δ15 desaturase protein from Neurospora crassa, both involved in the desaturation of endogenous fatty acids into stearidonic acid. The molecular characterisation of soybean MON does not raise safety issues. Soybean MON differs from the conventional counterpart in its fatty acid profile. The safety assessment of the newly expressed desaturases identified no concerns regarding potential toxicity and allergenicity. Nutritional assessment of soybean MON and derived food products did not identify concerns about human health and nutrition. Consumption of MON soybean oil replacing other oils in food is not expected to result in adverse effects from increased SDA intake as shown in different exposure scenarios. There are no indications of an increased likelihood of establishment and spread of feral soybean plants. Considering the scope of the application, potential interactions of soybean MON with the biotic and abiotic environment were not considered a relevant issue. Environmental risks associated with a theoretically possible horizontal gene transfer from soybean MON to bacteria have not been identified. The post-market environmental monitoring plan and reporting intervals are in line with the intended uses of soybean MON Since the use of oil derived from the soybean MON will result in a higher intake of SDA, a post-market monitoring plan is recommended to confirm the exposure assessment using realistic consumption data for the European population. European Food Safety Authority, 2014 KEY WORDS GMO, soybean (Glycine max (L). Merr.), MON 87769, stearidonic acid, Regulation (EC) No 1829/2003, gamma-linolenic acid, alpha-linolenic acid 1 On request from the Competent Authority of the UK on an application (EFSA-GMO-UK ) submitted by Monsanto, Question No EFSA-Q , adopted on 9 April Panel members: Salvatore Arpaia, Andrew Nicholas Edmund Birch, Andrew Chesson, Achim Gathmann, Jürgen Gropp, Patrick du Jardin, Lieve Herman, Hilde-Gunn Hoen-Sorteberg, Huw Jones, Sirpa Kärenlampi, Jozsef Kiss, Gijs Kleter, Martinus Lovik, Antoine Messéan, Hanspeter Naegeli, Kaare Magne Nielsen, Jaroslava Ovesna, Joe Perry, Nils Rostoks and Christoph Tebbe. Correspondence: gmo@efsa.europa.eu 3 Acknowledgement: The Panel wishes to thank the members of the Working Groups on Molecular Characterisation, Food and Feed Risk Assessment and Environmental Risk Assessment for the preparatory work on this scientific opinion; and EFSA staff: Hermann Broll, Antonio Fernandez Dumont, Sylvie Mestdagh and Matthew Ramon for the support provided to this scientific opinion. Suggested citation: EFSA Panel on Genetically Modified Organisms (GMO), Scientific Opinion on application (EFSA- GMO-UK ) for the placing on the market of soybean MON genetically modified to contain stearidonic acid, for food and feed uses, import and processing under Regulation (EC) No 1829/2003 from Monsanto. EFSA Journal 2014;12(5):3644, 41 pp. doi: /j.efsa Available online: European Food Safety Authority, 2014

2 SUMMARY Following the submission of an application (EFSA-GMO-UK ) under Regulation (EC) No 1829/2003 from Monsanto, the Panel on Genetically Modified Organisms (GMO) was asked to deliver a scientific opinion on soybean MON (unique identifier MON ), genetically modified (GM), to contain stearidonic acid, for food and feed uses, import and processing. In delivering its scientific opinion, the EFSA GMO Panel considered the application EFSA-GMO- UK , additional information supplied by the applicant, scientific comments submitted by Member States and relevant scientific publications. The scope of application EFSA-GMO-UK is for food and feed uses, import and processing of soybean MON and all derived products, but excludes cultivation in the European Union. The EFSA GMO Panel assessed soybean MON with reference to the intended uses and appropriate principles described in the Guidance Document of the Scientific Panel on Genetically Modified Organisms for the risk assessment of GM plants and derived food and feed (EFSA, 2006). The scientific assessment included molecular characterisation of the inserted DNA and the newly expressed proteins. A comparative analysis of agronomic traits and composition was undertaken, and the safety of the new protein and the whole food/feed were evaluated with respect to potential toxicity, allergenicity and nutritional quality. An assessment of environmental impacts and the post-market environmental monitoring plan were also undertaken. Meristematic tissue excised from the embryos of germinated seeds of conventional soybean A3525 was transformed using Agrobacterium tumefaciens and expresses the Primula juliae Δ6 desaturase gene and the Nc.Fad3 gene, which provides the expression of the Neurospora crassa Δ15 desaturase intended to modify the lipid profile of the extracted oil. The molecular characterisation data establish that the GM soybean MON contains a single insert consisting of the Pj.D6D and Nc.Fad3 expression cassettes. No other parts of the plasmid used for transformation could be detected in soybean MON Bioinformatic analyses and genetic stability studies did not raise safety issues. The levels of the PjΔ6D and NcΔ15D proteins in soybean MON have been sufficiently characterised to inform the subsequent assessment. A comparative analysis of soybean MON identified no phenotypic or agronomic differences with respect to its conventional counterpart (soybean A3525) and to non-gm soybean reference varieties. However, it confirmed that the composition of soybean MON differs from that of the conventional counterpart and non-gm soybean reference varieties. The newly expressed desaturases in soybean MON seeds resulted in an alteration of the fatty acid profile, leading to the appearance of four new fatty acids (stearidonic acid (SDA), -linolenic acid and two trans-fatty acids) and a reduction in linoleic acid (LA). The safety assessment identified no concerns regarding the potential toxicity and allergenicity of the newly introduced desaturase proteins. There are no indications that the genetic modification might change the overall allergenicity of soybean MON when compared with that of its conventional counterpart. The EFSA GMO Panel concludes that the estimated changes in fatty acid intake by consumers using oil from MON are unlikely to constitute a toxicological risk or to have negative nutritional consequences for humans. Based on the results of studies in rats, it is concluded that feeding stuffs derived from defatted soybean MON are as safe and nutritious as those derived from other non-gm soybean varieties. Based on different exposure scenarios, the EFSA GMO Panel concludes that the proposed use of MON soybean oil in foods is not expected to result in intakes of SDA with adverse effects and that the other changes in the dietary fatty acid pattern are unlikely to have negative nutritional consequences for humans. The EFSA GMO Panel notes that the quantitative dietary estimates described here would have to be revisited if the oil produced by soybean MON were to be extensively used in food products not considered in this assessment, for example as dietary supplements or to modify animal feed products. EFSA Journal 2014;12(5):3644 2

3 The EFSA GMO Panel recommends a post-market monitoring plan to confirm the exposure assessment using consumption data for the European populations. Considering the scope of application EFSA-GMO-UK , there is no requirement for a scientific assessment of possible environmental effects associated with the cultivation of this GM soybean. There are no indications of an increased likelihood of establishment and spread of feral soybean MON plants in the case of accidental release into the environment of viable GM soybean seeds. Owing to the scope of application EFSA-GMO-UK , and the low level of exposure to the environment, potential interactions of the GM plant with non-target organisms were not considered a relevant issue by the EFSA GMO Panel. The theoretically possible transfer of the recombinant genes from soybean MON to environmental bacteria does not raise a concern owing to the lack of both an efficient transfer mechanism and an identified selective advantage. The scope of the post-market environmental monitoring (PMEM) plan provided by the applicant and the reporting intervals are in line with the intended uses of soybean MON and the guidance document of the EFSA GMO Panel on PMEM of GM plants (EFSA, 2011). In addition, the EFSA GMO Panel acknowledges the approach proposed by the applicant to put in place appropriate management systems to restrict environmental exposure in cases of accidental release of viable seeds of soybean MON The EFSA GMO Panel agrees with the reporting intervals proposed by the applicant in its PMEM plan. In conclusion, the EFSA GMO Panel considers that the information available for soybean MON addresses the scientific issues indicated by the Guidance document of the EFSA GMO Panel and the scientific comments raised by the Member States, and that soybean MON is as safe as its conventional counterpart and is unlikely to have adverse effects on human and animal health and the environment in the context of the scope of this application. EFSA Journal 2014;12(5):3644 3

4 TABLE OF CONTENTS Abstract... 1 Summary... 2 Table of contents... 4 Background... 5 Terms of reference... 6 Assessment Introduction Issues raised by Member States Molecular characterisation Evaluation of relevant scientific data Transformation process and vector constructs Transgene constructs in the GM plant Information on the expression of the insert Inheritance and stability of inserted DNA Conclusion Comparative analysis Evaluation of relevant scientific data Choice of comparator and production of material for the comparative assessment Agronomic traits and GM phenotype Compositional analysis Conclusion Food/feed safety assessment Evaluation of relevant scientific data Effect of processing Toxicology Animal studies with the food/feed derived from GM plants Allergenicity Nutritional assessment of GM food/feed Post-market monitoring of GM food/feed Scientific correctness of proposed labelling Conclusion Environmental risk assessment and monitoring plan Evaluation of relevant scientific data Post-market environmental monitoring Conclusion Overall conclusions and recommendations Documentation provided to EFSA References EFSA Journal 2014;12(5):3644 4

5 BACKGROUND On 20 October 2009, the European Food Safety Authority (EFSA) received from the Competent Authority of United Kingdom (UK) an application (Reference EFSA-GMO-UK ), for authorisation of genetically modified (GM) soybean MON submitted by Monsanto Europe S.A/NV within the framework of Regulation (EC) No 1829/ on genetically modified food and feed for food and feed uses, import and processing. After receiving the application EFSA-GMO-UK and in accordance with Articles 5(2)(b) and 17(2)(b) of Regulation (EC) No 1829/2003, EFSA informed Member States and the European Commission, and made the summary of the application available to the public on the EFSA website. 5 EFSA initiated a formal review of the application to check compliance with the requirements laid down in Articles 5(3) and 17(3) of Regulation (EC) No 1829/2003. On 26 January 2010, EFSA received additional information (requested on 27 November 2009). On 16 February 2010, EFSA declared the application valid in accordance with Articles 6(1) and 18(1) of Regulation (EC) No 1829/2003. EFSA made the valid application available to Member States and the European Commission, and consulted nominated risk assessment bodies of Member States, including national Competent Authorities within the meaning of Directive 2001/18/EC 6 following the requirements of Articles 6(4) and 18(4) of Regulation (EC) No 1829/2003, to request their scientific opinion. Member States had three months after the date of receipt of the valid application (until 16 May) to make their opinion known. The scope defined by the applicant: includes all food and feed products containing, consisting or produced from soybean MON including products from inbreeds and hybrids obtained by conventional breeding of this soybean product. The application also covers the import and industrial processing of soybean MON for all potential uses as any other soybean. The EFSA GMO Panel carried out an evaluation of the scientific risk assessment of the GM soybean MON On 3 may 2010, 5 August 2010, 13 October 2010, 20 July 2011, 7 February 2012, 9 January 2013 and 10 October 2013, the EFSA GMO Panel requested additional information. The applicant provided the requested information on 14 June 2010, 1 October 2010, 24 February 2011, 10 November 2011, 20 September 2012, 18 February 2013, 21 May 2013 and 3 January 2014, respectively. In giving its scientific opinion to the European Commission, the Member States and the applicant, and in accordance with Articles 6(1) and 18(1) of Regulation (EC) No 1829/2003, EFSA has endeavoured to respect a time limit of six months from the acknowledgement of the valid application. As additional information was requested by the EFSA GMO Panel, the time limit of six months was extended accordingly, in line with Articles 6(1), 6(2), 18(1), and 18(2) of Regulation (EC) No 1829/2003. According to Regulation (EC) No 1829/2003, the EFSA opinion shall include a report describing the assessment of the food and feed and stating the reasons for its opinion and the information on which its opinion is based. This document is to be seen as the report requested under Articles 6(6) and 18(6) of that Regulation and thus will be part of the EFSA overall opinion in accordance with Articles 6(5) and 18(5). 4 Regulation (EC) No 1829/2003 of the European Parliament and of the Council of 22 September 2003 on genetically modified food and feed. OJ L 268, , p Available online: 6 Directive 2001/18/EC of the European Parliament and of the Council of 12 March 2001 on the deliberate release into the environment of genetically modified organisms and repealing Council Directive 90/220/EEC. OJ L 106, , p EFSA Journal 2014;12(5):3644 5

6 TERMS OF REFERENCE The EFSA GMO Panel was requested to carry out a scientific risk assessment of soybean MON for food and feed uses, import and processing in accordance with Articles 6(6) and 18(6) of Regulation (EC) No 1829/2003. Where applicable, any conditions or restrictions which should be imposed on the placing on the market and/or specific conditions or restrictions for use and handling, including post-market monitoring requirements based on the outcome of the risk assessment and, in the case of GMOs or food/feed containing or consisting of GMOs, conditions for the protection of particular ecosystems/environment and/or geographical areas should be indicated in accordance with Articles 6(5)(e) and 18(5)(e) of Regulation (EC) No 1829/2003. The EFSA GMO Panel was not requested to give a scientific opinion on information required under Annex II to the Cartagena Protocol. The EFSA GMO Panel did consider if there is a need for a specific labelling in accordance with Articles 13(2)(a) and 25(2)(c) of Regulation (EC) No 1829/2003. However, it did not consider proposals for methods of detection (including sampling and the identification of the specific transformation event in the food/feed and/or food/feed produced from it), which are matters related to risk management. EFSA Journal 2014;12(5):3644 6

7 ASSESSMENT 1. Introduction The scope of application EFSA-GMO-UK is for food and feed uses, import and processing of all derived products of soybean MON such as the production of refined oil and lecithin as food or food ingredients and the use of protein rich meal in animal feed. Soybean MON (unique identifier MON ) was assessed with respect to its scope, taking account of the requirements described in the applicable guidance documents (EFSA, 2006). The risk assessment presented here is based on the information provided in the application submitted in the European Union (EU), scientific comments raised by the Member States and relevant scientific publications. The genetic modification introduced in soybean MON results in the expression in the seeds of two novel desaturase enzymes intended to modify the lipid profile of the extracted oil. The first, a Δ15 desaturase, is active in the conversion of linoleic acid [(C18:2 (n-6)] (LA) to α-linolenic acid [C18:3 (n-3)] (ALA). The second enzyme, a Δ6 desaturase, promotes the conversion of ALA to octadecatetraenoic acid [C18:4 (n-3)], also known as stearidonic acid (SDA), which accumulates in the seed. The same enzyme also catalyses the conversion of LA to -linolenic acid [C18:3 (n-6)] (GLA), a precursor of arachidonic acid and the eicosanoids. SDA is a normal intermediate in the formation of the long chain omega-3 polyunsaturated fatty acids (PUFAs) eicosapentaenoic acid [(C20:5 (n-3)] (EPA) and docosahexaenoic acid [(C22:6 (n-3)] (DHA). However, in humans, the conversion of ALA to SDA is slow. Direct consumption of SDA avoids this step in the biosynthesis and can result in a more efficient synthesis of the higher chain-length PUFAs. Three major processed fractions are produced from whole soybean seeds: oil, protein rich meal and lecithin. The oil derived from soybean MON is intended to be identity preserved to maintain its value and assure its appropriate use in food applications. According to the applicant, it is foreseen to be added to foods as an ingredient that provides a precursor for EPA and DHA, in most cases replacing a portion of other oils in the diet. The SDA soybean oil is intended to be used only by the food industry and, according to the applicant, will not be available as home-use oil. 7 The high content of PUFAs makes it unsuitable for high temperature operations such as frying. 8 According to the applicant, the extracted protein rich meal will be used exclusively for animal feed. The applicant considers that, given the nature of the oil and separation from seeds of other soybean varieties needed to preserve identity, it is unlikely that whole soybean (full fat) or refined oil from MON would be used in animal diets. Nevertheless, the scope of the present application implies that soybean MON may be treated as any other soybean and this possibility is assessed below. 2. Issues raised by Member States The comments raised by the Member States are addressed in Annex G of the EFSA overall opinion and were taken into consideration during the evaluation of the risk assessment. 9 7 Technical dossier p Technical dossier p Available online: EFSA Journal 2014;12(5):3644 7

8 3. Molecular characterisation 3.1. Evaluation of relevant scientific data Transformation process and vector constructs Meristematic tissue excised from the embryos of germinated seeds of conventional soybean A3525 was transformed with the binary plasmid PV-GMPQ1972 using Agrobacterium tumefaciens (also known as Rhizobium radiobacter) strain ABI. The plasmid PV-GMPQ1972 contained two T-DNAs. T-DNA I contained the Pj.D6D gene expression cassette providing the expression of Primula juliae Δ6 desaturase and the Nc.Fad3 gene expression cassette, which provides the expression of the Neurospora crassa Δ15 desaturase. T-DNA II contained the CP4 epsps cassette conferring tolerance to glyphosate, which served as a selectable marker for transformation. 10 The two-t-dna system enabled the cassettes encoding the traits of interest and the selectable marker to be inserted at two independent genetic loci within the genome of the soybean. After self-pollination of the transformed R0 plant, an R1 plant (designated as MON 87769) that contained a single T-DNA I, but not T-DNA II, was selected for further development. The two T-DNAs present in plasmid PV-GMPQ1972 consisted of the following elements between their respective right and left border regions: T-DNA I (Pj.D6D and Nc.Fad3 expression cassettes): seed-specific promoter and leader sequence (P-7Sα ) of soybean Sphas1 gene (Sphas1 gene encodes β-conglycinin, a 7Sα seed storage protein) to direct the transcription in the seed; coding sequence for the fatty acid Δ6 desaturase from P. juliae (primrose) (Pj.D6D); 3 non-translated region of the tml gene from the A. tumefaciens octopine-type Ti plasmid that directs polyadenylation of the mrna (Ttml); promoter and leader sequence (P-7Sα) of the soybean Sphas2 gene (Sphas2 gene encodes the α-subunit of β-conglycinin) to direct the transcription in the seed; coding sequence for the fatty acid Δ15 desaturase from N. crassa (Nc.Fad3); 3 non-translated region of the Pisum sativum (garden pea) rbcs2 gene that directs polyadenylation of the mrna (rbcs2 gene encodes Rubisco small subunit) (T-E9). The Nc.Fad3 protein of N. crassa contained a single amino acid change (from threonine to alanine at the first amino acid after the start codon). T-DNA II (CP4 epsps expression cassette): FMV promoter (P-FMV) from Figwort mosaic virus 35S RNA gene, which drives transcription in most plant cells; 5 non-translated leader sequence from the Arabidopsis shkg gene (shkg gene encodes EPSPS) to enhance expression (L-ShkG); sequence encoding the transit peptide region of A. thaliana EPSPS to direct the CP4 EPSPS protein to the chloroplast (TS-CPT2); modified coding sequence of the aroa gene from Agrobacterium sp. strain CP4, encoding the EPSPS protein, to confer tolerance to glyphosate during the selection of transformants (CS-CP4 epsps); 3 non-translated region of the P. sativum rbcs2 gene that directs polyadenylation of the mrna (T-E9). Additional functional elements in the plasmid vector outside the T-DNAs, and thus not expected to be transferred to the soybean genome, were: oriv origin of replication to maintain the plasmid in Agrobacterium; ori-pbr322 origin of replication to maintain the plasmid in Escherichia coli; rop repressor of primer (ROP) protein to maintain plasmid copy number in E. coli; aada bacterial selectable marker (promoter and coding regions) to confer spectinomycin/streptomycin resistance Transgene constructs in the GM plant The DNA sequences inserted in the MON event were characterised by Southern analysis and by polymerase chain reaction (PCR) amplification of both the insert and the flanking regions Dossier: Part I Section C1, C2. 11 Dossier: Part I Section C3. 12 Dossier: Part I Section D2. EFSA Journal 2014;12(5):3644 8

9 Southern analyses indicated that soybean MON contains a single insert with one copy of the intact Pj.D6D and Nc.Fad3 expression cassettes. The insert and copy number were confirmed by multiple restriction enzyme/probe combinations covering the T-DNA region and flanking regions. No signal was observed with the overlapping probes corresponding to the PV-GMPQ1972 vector backbone and T-DNA II (except for the border sequences identical between the two T-DNAs). Some probes detected endogenous soybean sequences, as parts of the T-DNA I and T-DNA II cassettes were of soybean origin. The nucleotide sequence of the entire insert, as well as approximately 1 kb of both 5 and 3 flanking regions (933 and 831 bp, respectively), were determined from soybean MON The sequence of the insert confirmed the conclusions drawn from the Southern analyses. The insert is identical to the T-DNA I of PV-GMPQ1972, except for the deletion of 313 bp of the right border and the deletion of 168 bp of the left border region. The possible interruption of known endogenous soybean genes by the insertion of event MON was evaluated by bioinformatic analyses of the pre-insertion locus and of the genomic sequences flanking the insert. Comparison of the sequences of the flanking regions in MON to those in the parental soybean A3525 indicated that in MON a 9 bp DNA segment of endogenous DNA has been deleted and two (17 and 8 bp) filler DNAs were introduced immediately 5 and 3 to the insertion site, respectively. BLASTN searches were performed against the GenBank EST (Expressed Sequence Tag) and non-redundant nucleotide database and BLASTX search against the GenBank non-redundant amino acid database. These bioinformatic analyses did not reveal the interruption of any known endogenous gene in the MON flanking regions. 13 The results of segregation (see Section 3.1.4) and bioinformatic analyses established that the insert is located in the nuclear genome. 14 In order to assess whether the open reading frames (ORFs) present within the insert and spanning the junction sites raise any safety issue, their putative translation products were compared with databases for similarities to known allergens and toxins using suitable algorithms. No significant similarities were found Information on the expression of the insert The levels of the integral membrane proteins PjΔ6D and NcΔ15D were estimated by semi-quantitative Western analysis with peptide antibodies developed against the soluble portion of the corresponding proteins. Data were analysed from replicated field trials in the USA across five locations in 2006 (n = 14) and five locations in 2007 (n = 15). As expected, since the newly expressed genes are under the control of seed-specific promoters, none of these proteins was detected in the leaf or root. In the 2006 growing season, the mean levels of PjΔ6D were estimated to be 16 μg/g dry weight (dw) (SD = 9.5 μg/g) with a range of μg/g dw in forage (aerial plant parts including immature seeds), 100 μg/g dw (SD = 63 μg/g) with a range of μg/g dw in immature seed and 1.8 μg/g dw (SD = 0.95 μg/g) with a range of μg/g dw in mature seed. The NcΔ15D levels were estimated to be 14 μg/g dw (SD = 6.8 μg/g) with a range of μg/g dw in forage, 200 μg/g dw (SD = 89) with a range of μg/g dw in immature seed and 10 μg/g dw (SD = 6.5) with a range of μg/g dw in mature seed. Similar levels were observed in the samples from The levels of both proteins were markedly higher in immature seeds than in mature seeds for the two seasons and the highest levels were observed in the 2006 growing season. The immature seed was used as a source of both proteins for the toxicological assessment (Section ) Inheritance and stability of inserted DNA The integration of the insert in the nuclear genome was confirmed by Southern analysis and PCR. Stability of the inserted DNA was studied by Southern analysis from four consecutive generations, all 13 Dossier: Part I Section D2. Additional information: 21/5/ Dossier: Part I Section D2, D5. Additional information: 21/5/ Additional information: 21/5/ Dossier: Part I Section D3. EFSA Journal 2014;12(5):3644 9

10 of them produced by self-pollination (R3 to R6). The insert was stable and followed the Mendelian inheritance pattern of a single locus. 17 Phenotypic stability was indicated by analysing the presence of the T-tml 3 genetic element (by quantitative structure-specific endonuclease-based assay) over three generations produced by self-pollination after an initial backcross of MON (R4 generation) with a conventional soybean variety Conclusion The molecular characterisation data provided by the applicant establish that soybean MON contains a single insertion consisting of two intact expression cassettes (Pj.D6D and Nc.Fad3). No other parts of the plasmid used for transformation are present in the transformed plant. Bioinformatic analysis of the 5 and 3 flanking regions did not reveal disruption of known endogenous genes or regulatory sequences, or creation of ORFs that would cause a safety issue. The stability of the inserted DNA was confirmed over several generations and a Mendelian inheritance pattern was demonstrated. The EFSA GMO Panel concludes that the molecular characterisation does not raise safety issues. The levels of the PjΔ6D and NcΔ15D proteins in soybean MON 87769, have been sufficiently characterised to inform the subsequent assessment. 4. Comparative analysis 4.1. Evaluation of relevant scientific data Choice of comparator and production of material for the comparative assessment Data on agronomic and phenotypic characteristics of soybean MON 87769, its conventional counterpart and a set of non-gm commercial varieties were collected in field trials performed in the USA in 2006 and These field trials also supplied seed and forage material for compositional analysis of the various soybean materials. The composition of soybean MON was compared with that of the conventional counterpart Asgrow variety A3525, which was the commercial soybean variety originally used to establish transformation event MON In both years, the field trial was carried out at five geographical sites representative of the soybean cultivation areas of the USA. 19 At each site, soybean MON 87769, the conventional counterpart and three non-gm commercial varieties were planted following a randomised complete block design with three replicates. All the soybeans at each field trial site were grown under normal agronomic management for that geographical region. In total, 10 different commercial non-gm soybean varieties were included in 2006 and 15 in These were cultivated to provide data on the natural variation in agronomic and phenotypic characteristics and composition amongst commercial soybean varieties. However, when an event-specific PCR analysis was made, one of the replicates in one of the reference lines from one of the field trials in 2007 was identified as contaminated with MON and excluded from the analysis. The reference material 20 was used to estimate a range of baseline values that are common to commercial soybean varieties for each parameter studied. On request from the EFSA GMO Panel, the applicant provided ranges of baseline values based only on non-gm soybean reference varieties. 21 All materials were grown at normal agronomic conditions for the specific geographical region. One of the field trial sites in 2006 was excluded from the analysis because two of the three control plots had poor soybean stands owing to a malfunction of the planting equipment. 17 Dossier: Part I Section D5. 18 Technical Dossier/Section D7.1 i) and ii). 19 The field trials in 2006 were performed at two sites in Iowa, and one site each in Illinois, Michigan and Ohio, and in 2007 at one site each in Iowa, Michigan, Nebraska, Pennsylvania and Wisconsin. 20 In 2006 the reference lines were: A3244, ST3600, Stewart SB3454, DKB34-51, ST3608, Pioneer 93M50, Pioneer 93B82, Lewis 372, AG3505, CST3461 (STS), ST3300, CST37002, ST3870, A2869, ST2788, Lewis 392, A2804, and A2553. In 2007 they were: DKB34-51, Hoegemeyer 333, CST3461 (STS), ST3600, AG3505, ST3300, Stewart SB3454, CST37002, Pioneer 93M50, Midland 363, A3244 and ST3608 (reference lines in bold are GM soybeans). 21 Additional information: 30/9/2010. EFSA Journal 2014;12(5):

11 Agronomic traits and GM phenotype Scientific Opinion on genetically modified soybean MON The phenotypic and agronomic characteristics evaluated were early stand count, seedling vigour, plant growth stages, days to 50 % flowering, flower colour, plant pubescence, plant height, lodging, pod shattering, final stand count, seed moisture, 100-seed weight, test weight (g/250 ml) and yield. During each year, soybean MON was compared with soybean A3525 within each site and across sites. In the phenotypic comparison none of the parameters differed between soybean MON and the conventional counterpart in the statistical analysis across sites. In the individual site analysis, a total of 23 statistically significant differences were detected out of 204 comparisons. However, the mean values observed for soybean MON fell within the minimum and maximum mean values estimated for the reference lines. Therefore, the GMO Panel did not consider that these differences would require further assessment in the context of the scope of this application. No developmental differences (flower colour, plant pubescence and plant growth stage data) or altered pollen parameters (pollen diameter, morphology and viability) were observed between soybean MON and its conventional counterpart Compositional analysis Soybean seeds were harvested from the field trials in the USA in 2006 and 2007, and analysed for proximates (protein, fat, ash and moisture), fibre fractions (acid detergent fibre (ADF) and neutral detergent fibre (NDF)), amino acids, fatty acids, vitamin E, anti-nutrients (i.e. phytic acid, trypsin inhibitor, lectins, stachyose and raffinose) and other secondary metabolites (isoflavones). Forage was analysed for proximates and for ADF and NDF. The 75 parameters analysed (68 in seeds and 7 in forage) were those recommended by OECD (2001) with the addition of extra fatty acids. Values for 26 endpoints frequently were at levels below the limit of quantification. When this occurred in more than 50 % of the samples, the analyte was omitted from the analysis. 22 Comparison of forage parameters showed no statistically significant differences across locations in either season. Statistically significant differences between the seed fatty acid composition of soybean MON and its conventional counterpart were observed as expected owing to the genetic modification. However, the total fat content of the seeds did not differ between soybean MON and its conventional counterpart. As shown in Table 1, the fatty acid composition in both years was qualitatively similar, although small differences in the proportions of the various fatty acids among years and sites were observed. In both years the most notable changes were a reduction in linoleic acid from % to % and in oleic acid from % to % of total fatty acids. This reduction was accompanied by the appearance of the two metabolites SDA ( %) and GLA ( %). In addition, low amounts of two trans-fatty acids previously not found in measurable concentrations in soybean oil, 9c,12c,15t trans-ala (18:3) at % and 6c,9c,12c,15t trans-sda (C18:4) at %, were detected. The statistical analysis also revealed increased protein and reduced carbohydrate content in seeds. In agreement with this observation, the level of 17 of the 18 amino acids analysed was significantly increased in 2006 and the level 7 of the 18 increased in the following year. Differences in the levels of all amino acids were always within the variation defined by the soybean reference varieties included in the field trials. As the carbohydrate content is calculated by taking the difference from the sum of the other proximate constituents, the apparent reduction of this parameter is a consequence of the altered 22 Components excluded from the compositional analysis owing to predominant observations below the limit of quantitation of the assay were: 8:0 caprylic acid, 10:0 capric acid, 12:0 lauric acid, 14:0 myristic acid, 14:1 myristoleic acid, 15:0 pentadecylic acid, 15:1 pentadecenoic acid, 16:1 palmitoleic acid, 17:0 margaric acid, 17:1 heptadecylenic acid, 18:1 total trans fatty acids, 18:2 6c,9c isolinoleic acid, 18:2 total trans fatty acids, 18:3 gamma linolenic acid, 18:3 other 18:3 trans fatty acids, 18:3 9c,12c,15t trans ALA, 18:4 stearidonic acid (SDA), 18:4 6c,9c,12c,15t trans SDA, 20:2 11c,14c eicosadienoic acid, 20:3 11c,14c,17c eicosatrienoic acid, 20:4 arachidonic acid, 20:5 eicosapentaenoic acid (EPA), 22:1 erucic acid, 22:5 docosapentaenoic acid (DPA), 22:6 docosahexaenoic acid (DHA) and 24:0 lignoceric acid. EFSA Journal 2014;12(5):

12 protein content. The EFSA GMO Panel identified no biological relevance in the observed altered protein and carbohydrate content of the soybean seeds requiring further assessment. EFSA Journal 2014;12(5):

13 Table 1: Fatty acid profile (% of total fatty acids; mean and range) in seeds and in refined bleached deodorised oil produced from soybean MON and its conventional counterpart (A3525) Compound Fatty acid composition (% of total fatty acids) in MON soybean seeds 2006 USA (five locations) 2007 USA (five locations) 2006 USA (two locations) Reference MON A3525 MON A3525 MON RBD processed oil A3525 RBD processed oil values* Myristic acid (C14:0) Mean n.d. n.d. n.d. n.d n.g. Range n.d. n.d. n.d. n.d n.d. Palmitic acid (C16:0) Mean Range Palmitoleic acid (C16:1) Mean n.d. n.d. n.d. n.d n.g. Range n.d. n.d. n.d. n.d n.d. Heptadecanoic acid Mean n.d. n.d. n.d. n.d n.g. (C17:0) Range n.d. n.d. n.d. n.d n.d. Stearic acid (C18:0) Mean Range Oleic acid (C18:1) Mean Range c,9c Isolinoleic acid Mean n.d. n.d. n.d. n.d n.g. (C18:2) Range n.d. n.d. n.d. n.d n.d. Linoleic acid (C18:2) Mean Range α-linolenic acid (C18:3) Mean Range Trans- -Linolenic acid Mean 0.44 n.d n.d n.g. (C18:3) 9c,12c,15t Range n.d n.d n.d. Other trans-linolenic Mean n.d. n.d. n.d. n.d n.g. acids (C18:3)** Range n.d. n.d. n.d. n.d n.d. -Linolenic acid (C18:3) Mean 7.09 n.d n.d n.d. n.g. Range n.d n.d n.d. n.d. Stearidonic acid (C18:4) Mean n.d n.d n.d. n.g. Range n.d n.d n.d. n.d. Trans-Stearidonic acid Mean 0.18 n.d n.d n.d. n.g. (C18:4) 6c,9c,12c,15t Range n.d n.d n.d. n.d. Arachidic acid (C20:0) Mean EFSA Journal 2014;12(5):

14 Compound Fatty acid composition (% of total fatty acids) in MON soybean seeds 2006 USA (five locations) 2007 USA (five locations) 2006 USA (two locations) Reference MON A3525 MON A3525 MON RBD processed oil A3525 RBD processed oil values* Range Eicosenoic acid (C20:1) Mean n.g. 11c Range Behenic acid (C22:0) Mean Range Lignoceric acid (C24:0) Mean n.d. n.d. n.d. n.d n.g. Range n.d. n.d. n.d. n.d n.d. Statistically significant differences (p = 0.05) and newly appearing compounds are shown on a shaded background. n.g. = not given; n.d. = not detectable; RBD = refined, bleached and deodorised. *Bold in this column: data from ILSI (2008), whereas non-bold data refer to reference lines in field trials in the USA in 2006 and **The other trans-(c18:3) linolenic acids were 9c,12t,15c-; 9t,12c,15t-; and 9t,12c,15-(C18:3) linolenic acid. EFSA Journal 2014;12(5):

15 A higher vitamin E content in soybean MON was observed at only one of the five sites in 2007, but not in Of the anti-nutrients, the phytic acid level was increased in soybean MON in However, the levels were within the 99 % tolerance interval of the levels of the reference soybean varieties. A reduction in daidzein and genistein content (30 36 %) in 2006 and 2007 and in glycitein level (only in 2006) were also observed. 23 However, these reduced isoflavone levels were within the 99 % tolerance interval of the reference soybean varieties included in the field trials. There were several additional statistically significant differences identified in the per location statistical analysis, but the levels observed in soybean MON were within the range of values observed in the reference varieties and therefore did not raise concern Conclusion A comparison of soybean MON with its conventional counterpart (soybean A3525) and non- GM soybean reference varieties identified no phenotypic or agronomic differences requiring further assessment. The newly expressed desaturases in soybean MON seeds resulted in an alteration of the fatty acid profile; this alteration is characterised by the appearance of four new fatty acids (SDA, GLA and two trans-fatty acids) coupled with the reduction in LA. The safety and nutritional impacts of the altered fatty acid levels are evaluated in Section 5. The EFSA GMO Panel identified no biological relevance in the other observed differences which, therefore, do not require further assessment. 5. Food/feed safety assessment 5.1. Evaluation of relevant scientific data Effect of processing Soybeans were harvested from two of the five sites in the USA in in order to perform compositional analyses on processed fractions, including defatted and toasted meal; refined, bleached and deodorised oil; protein isolate; and crude lecithin, derived from MON 87769, A3525 and eight conventional reference soybean varieties. The soybean meal was analysed for proximates, fibre fractions, amino acids, fatty acids, phytic acid and trypsin inhibitors, the soybean oil for fatty acids and vitamin E, the protein isolate for amino acids, fatty acids and moisture and, finally, crude lecithin for fatty acids and phosphatides. In all cases where a fatty acid analyses was made, the compounds analysed were the extended battery of fatty acids defined in Section In total, 129 analytes were determined in the compositional comparison of soybean products. Comparing the defatted and toasted meal 25 produced from soybean MON with similar meals from the conventional counterpart showed changes in the fatty acid profile that mirrored the differences seen with the whole soybean (i.e. reduced LA and oleic acid and the appearance of SDA and GLA). Statistically significant changes in the concentration of six other constituents were also found. These were slight increases in the level of four amino acids and in ADF in soybean MON and a reduction in the calculated carbohydrate content. The fatty acid changes were expected owing to the genetic modification. The small increase in amino acid content and the reduced carbohydrate content of the meal mirrored the differences observed in the whole seed as would be expected. These changes have been previously considered (see Section ) and found to be of no biological relevance. As expected, refined, bleached and deodorised oil from soybeans MON differs from that of the corresponding oil produced from A3523 soybean. When the compositional data on processed oils from both types of soybean were compared (Table 1), statistically increased levels of palmitic acid, stearic acid, trans-ala and vitamin E were observed, whereas the level of lignoceric acid was reduced. The 23 Technical Dossier/Section D7. 24 Technical Dossier/Section D7.1 iii). 25 Frequently containing around 1 % fat (not more than 3 %) compared with 20 % in full fat meal. EFSA Journal 2014;12(5):

16 level of LA was extensively reduced (from % in the conventional counterpart to % of the fatty acids in soybean MON 87769). In addition to these changes, three of the new fatty acids identified in the whole seed were also seen in the refined oil from MON (SDA, GLA and trans-sda, constituting %, % and % of the total fatty acids respectively). Small quantities of trans-ala were present in all types of refined, bleached and deodorised soybean oil, suggesting that small quantities of this trans-fatty acid may be produced during processing of the oil. Owing to the lack of commercially available standards for 9c,12t,15c, 9t,12c,15t and 9t,12c,15c C18:3 trans-fatty acids, these could not be individually quantified. 26 However, the sum of these other C18:3 trans-fatty acids was similar in processed oil of soybean MON ( %), soybean A3525 ( %) and reference soybeans ( %). The total trans-fatty acid levels in commercial grade SDA soybean oil was reported to range from 0.5 % to 0.8 % of total fatty acids. Besides the reduced level of linoleic acid (C18:2), and the new fatty acids in oil produced from soybean MON 87769, changes in the level of the various fatty acids (and vitamin E) normally found in conventional soybean oil were small, and were within the 99 % tolerance interval of the level of the various fatty acids defined by oil produced from the conventional reference soybean varieties included in the study. On request from the EFSA GMO Panel, the applicant supplied information on the oxidative stability of the SDA enriched oil obtained from soybeans MON At room temperature (25 ºC in air) the oil maintains acceptable quality for at least 72 days, while under accelerated ageing conditions (55 ºC in air) it is substantially shorter; however, it kept an acceptable quality for at least four to five days. These data indicate that the transition times for SDA soybean oil at the accelerated ageing conditions are within the ranges observed for other omega-3 oils, such as stabilised fish and algal oils. Storage of the SDA soybean oil under nitrogen at room temperature maintained the quality of the oil for at least nine months (FDA, 2009). A comparison of the composition of protein isolates from soybean MON and soybean A3525 revealed a slightly reduced level of leucine, which was within the 99 % tolerance interval of the conventional reference soybean varieties. No other significant differences were observed other than those related to the lipid content. Protein isolate is derived from defatted soy flour and therefore contains even lower levels of lipids, typically 3 % total fat. As expected, the pattern of individual lipids found reflected the starting material: LA in protein isolate from soybean MON was reduced, and trans-ala and ALA increased. The fat phase of the protein isolate produced from soybean MON also contained SDA, GLA and trans-sda. Comparing the composition of crude lecithin produced from MON and soybeans A3525 harvested across sites revealed no difference in the concentration of the four phosphatides investigated. Whereas the level of arachidic acid (C20:0) was increased, the level of lignoceric acid (C24:0) was reduced, but the levels of these fatty acids where within the 99 % tolerance interval of the conventional reference soybean varieties. The level of linoleic acid (C18:2) was reduced from % of total fatty acids in soybean A3525 to % of total fatty acids in soybean MON The crude lecithin derived from soybean MON contained SDA, GLA and trans- SDA, which are usually not detected in lecithin from conventional soybeans. In conclusion, the comparative compositional analyses of products derived from soybean MON 87769, including defatted and toasted meal; refined, bleached and deodorised oil; protein isolate; and crude lecithin, identified that, besides the expected changes in fatty acid composition, levels of other analysed constituents in soybean MON either were comparable with those in the conventional counterpart (soybean A3525) or, when significantly altered, were within the range of that particular constituent observed in products processed from the reference soybean varieties. 26 Additional information: 23/2/2011. EFSA Journal 2014;12(5):

17 Toxicology Scientific Opinion on genetically modified soybean MON This assessment concentrates on the newly expressed proteins Primula juliae Δ6 desaturase (PjΔ6D) and Neurospora crassa Δ15 desaturase (NcΔ15D) and on the four fatty acids stearidonic acid (C18:4; SDA), γ-linolenic acid (C18:3; GLA), 9c,12c,15t trans-ala (C18:3) and 6c,9c,12c,15t trans-sda (C18:4) produced in seeds of soybean MON normally not present at detectable levels in non- GM soybean seeds Proteins used for safety testing 27 The newly expressed PjΔ6D and NcΔ15D proteins used for safety testing were extracted from immature soybean MON seeds by solubilisation from membranes and subsequent purification by chromatographic procedures. The PjΔ6D protein in the extract had a concentration of 0.52 mg/ml, a purity of 47 % and an apparent molecular weight of 45.9 kda as determined by sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE). The 15 N-terminal amino acids of the PjΔ6D protein were shown to correspond to the amino acids deduced to constitute the coding region of the PjΔ6D gene present in soybean MON 87769, except for the terminal methionine, which is absent from the purified enzyme. It is known that post-translational modification of proteins frequently removes N-terminal methionine residues (Bradshaw et al., 1998). Matrix-assisted laser desorption ionisation time-of-flight (MALDI- TOF) analysis confirmed the identity of the MON produced PjΔ6D. A glycosylation assay identified no glycosylated proteins with a size around 46 kda. The NcΔ15D protein in the extract had a concentration of 0.62 mg/ml, a purity of 74 % and an apparent molecular weight of 46.2 kda as determined by SDS-PAGE. Also in this case the 15 N- terminal amino acids of the NcΔ15D protein corresponded to the amino acids deduced from the coding region of the MON soybean NcΔ15D gene, except that the terminal methionine was again absent in the purified enzyme. MALDI-TOF analysis confirmed the identity of the MON produced NcΔ15D. A glycosylation assay identified no glycosylated proteins with a size around 46 kda. The functional activities of both enzymes were investigated using a Saccharomyces cerevisiae PjΔ6 or NcΔ15 desaturase-expressing system. In vitro functional activity was also demonstrated for PjΔ6 desaturase, demonstrating 14 C-labelled stearidonic acid synthesis from 14 C-ALA-CoA incubated with crude homogenates of young, fresh MON seeds. For methodological reasons (lack of sensitivity of the method) it was not possible to demonstrate the functional activity of the NcΔ15 desaturase in soybeans MON The two protein preparations were used in in vitro degradation studies, in heat denaturation studies and in acute toxicity studies in mice Assessment of newly expressed proteins 28,29 The newly expressed proteins encoded by the desaturase genes in soybean MON naturally occur in the flowering plant Primula juliae and the ascomycete fungus Neurospora crassa. Only Neurospora crassa is used for food preparation in some regions of the world. Onchom is a soybeanbased press cake with N. crassa frequently consumed in Indonesia (Matsuo, 1997). In Brazil, N. crassa is used to process cassava in preparing a fermented drink (Park et al., 1982), and in France N. crassa is used in cheese production (Perkins and Davis, 2000). The PjΔ6D and NcΔ15D proteins expressed in soybean MON are homologous to desaturase proteins universally present in the human diet. Bioinformatics-supported searches in databases showed that the amino acid sequence of the PjΔ6D protein shares partial identity with other front-end 27 Technical Dossier/Section 7.8.1a. 28 Technical Dossier/Section 7.8.1b. 29 Additional information: 23/12/2013. EFSA Journal 2014;12(5):